Web Service Caching Issue

Hi All,

We are using a web service connector as our auth source. Recently, HR has made a change to a few of the mappings and we have made the matching modifications in our API call and mappings.

However, when running an aggregation I can see in the error message the mappings that have been deleted. I have checked in both vscode and via postman and can confirm the mapping doesn’t exist, despite the error message continually saying it does. I have also deleted the aggregation calls completely and the error message is the same. I also modified and deleted the test connection to get it to fail and it still connects successfully.

I have tried restarting the VAs. I cannot delete this source since it is the auth source and I can confirm that this issue is only occurring with this specific source as a separate web service connector is working as it should. Any suggestions on how to fix this would be great, we are in the middle of testing for a go live so having to blow up the auth source will very negatively impact the project.

3 Likes

hey @rgandhi , have you updated Attribute sync? All the transforms taht reference this “deleted” attributes?Or even IdentityProfile mappings.
try checking those and get back to us.

Hi @ipobeidi the account attribute that was using the old mapping was changed months ago to point to the new mapping from the HR source. Recently the HRIS team just deleted those old fields completely but we weren’t actually using them, they were just still included in our API call. There is no attribute sync for those fields and the transforms that reference anything just reference the account attribute, which remains the same.

The same change was made in prod and there is no issue there, it seems to just specifically be this source in sandbox.

Hi @rgandhi,

Have you enabled the logging and have seen that before operation and after operation rule the response json is being mapped correctly or there is some issue in the after-operation rule.

If possible, try enabling logger levels and also add one simple after operation to check the final response that is being generated

Thanks

I had similar issue in the past, this may be due to caching in the cloud (not the caching on VA). Here is what I did to get around issue -

  • On connector configuration page, temporarily change password to wrong one and do Test Connection, basically forcing it to fail. This will force cloud cache to clear.
  • Back on connector configuration page, update password back to correct one and do Test Connection.
    This hopefully fixes you issue.